The reason that potato plants form tubers, or the process of tuberization, has long puzzled both farmers and scientists. After all, related species such as the tomatoes do not form tubers. Of course we can say that potato contains the genes both to form the tubers and to regulate when these tubers form, but that begs the question as to the exact nature of the controls on the process. We know that tuberization is a regulated process, and in the wild relatives of potatoes and in some cultivars this is regulated by the length of day, or photoperiod. Thus tuberization occurs under short day lengths, and is also promoted by low night temperatures as would occur in the fall, because in nature the tuber is the plant's way of surviving from season to season. In most commercial varieties, this propensity to tuberize only in short days has largely been selected out, so that tuber formation takes place even under the long days of midsummer. From a scientist's point of view, however, this regulation of tuberization by short days in andigena-type potatoes provides a useful switch to ask what internal signals cause potatoes to produce tubers when at other times they do not. Most research on tuberization is done on such photoperiod-sensitive potatoes.
The first thing we need to note is that tuberization is not under the control of a single signal. When a large population of offspring from crosses of potatoes of differing genetic makeup, and degree or time of tuberization were analyzed genetically by Jan van den Berg, together with Elmer Ewing, it was found that tuberization was regulated by 12 locations on eight of the 12 chromosomes in potatoes, and if all six of the alleles (form of genes) favoring tuberization were present they could account for 98 percent of the promotion of tuberization. Thus tuberization is under the control of several genes, indicating the likelihood that tuberization is regulated by a balance of several factors.
One of these factors is clearly a group of hormones called gibberellins. These are called gibberellins because they were first discovered 80 years ago in rice infected with a disease fungus called Gibberella, which stimulated ultra-tallness (and also lack of grain) in the infected plants. When the fungus was grown in culture, it secreted the growth-active agent into the medium, which would then produce exceptionally tall growth when painted onto plants.
About 50 years ago it was realized that gibberellins were native plant hormones regulating many aspects of plant growth including tallness. Not only that, but there are many gibberellins, all based on a common chemical skeleton and differing in minor chemical modifications, which, however, can have profound differences on their action. Many chemical steps are needed for the plant to make growth-active gibberellin, and the last seven steps involve the final modifications to the gibberellin chemical skeleton, which is why any plant contains several gibberellins. These steps are under precise control, regulated both by the genetic makeup of the plant and the environment, especially day-length, as sensed by the plant.
A notable example of the regulation by gibberellins is that of height in peas. One hundred and fifty years ago, the father of genetics, the Austrian monk Gregor Mendel, demonstrated that height in peas was controlled by a single gene. Ten years ago, an Australian group including myself finally succeeded in showing that Mendel's tallness gene in fact encoded an enzyme involved in the final step in the creation of the growth-active gibberellin, named GA1, from the non-growth-active gibberellin GA20. (Gibberellins, now over 125, are numbered in the order of their discovery.) Thus tall peas have 20 times the content of growth-active gibberellins as compared to dwarf peas.
But back to potatoes: It has been known for quite a time that tuberization can be prevented by spraying potatoes with gibberellin (usually the gibberellin commercially available from fungus cultures, GA3). Thus active gibberellin is a tuberization inhibitor. In the early 1990s Jan van den Berg and I succeeded in identifying all the gibberellins in potatoes. Luckily they turned out to be the same as in peas and several other plants, so that we then knew the pathway of gibberellin biosynthesis in potatoes. In the andigena potatoes, the normal-height plants require short days for tuberization, whereas a dwarf mutant of an otherwise similar genetic background tuberizes under long days (see Figure 1). The dwarf plants contained less growth-active GA1, supporting the idea that tuberization was inhibited by gibberellins. We also examined the metabolism of the first gibberellin (GA12) along the biosynthetic pathway by supplying radioactive-labeled GA12 to the shoots. (Following a radioactive label, in this case radioactive carbon, from compound to compound, is a common way of analyzing metabolic pathways in living organisms.) Dwarf plants showed a reduction in the conversion in an early step in the pathway, so accounting for the fact that dwarf plants had a lower content of the final active gibberellin.
As another approach we tried feeding radioactive GA12 to potatoes that could or could not detect day length because of the presence or absence of the pigment (called phytochrome) needed to perceive photoperiod. Plants lacking the pigment to perceive photoperiod tuberize regardless of day length. We were unable to measure differences in GA1 production at the last stage of the pathway because the amounts were too small to detect. However, the plants lacking phytochrome not only had a peculiar morphology, with the leaf stalks extending perpendicularly upward, but their stems were longer, a sign normally indicating higher active gibberellin levels. This seemed backward from what was expected, namely that lower gibberellins should be associated with tuberization, and was puzzling.
So some more background: gibberellin biosynthesis occurs in the shoot, but, of course, tuber production occurs at the tip of the stolons.
The potato plants vigorously produce stolons as branches in the axils of the lower leaves. These will be above ground in a plant growing on flat ground, but below ground in ridged plants. The stolons spread the plant, and in this mode they are thin with a lot of extension growth. When stolons change to tuber formation, extension growth at the tip ceases (associated with a drop in growth-active gibberellin), and the tip expands laterally to form the tuber.
The answer to our puzzle came from a Spanish researcher, Salome Pratt. Elmer Ewing had visited her institute and talked about our gibberellin/photoperiod work, and she was intrigued. From experiments she finally worked out the following scheme: what is happening with regard to gibberellins in the leaves may not be the same as in the far distant (and underground) stolon tips.
While a drop in gibberellins everywhere, as in the dwarf plants, results in tuberization, the shoot may have an entirely different gibberellin profile compared to the stolons where tuberization is occurring.
Regardless of day length, gibberellin biosynthesis is actively occurring in the shoot because of high activity of a biosynthetic enzyme early in the pathway (named GA 20-oxidase).
However, in short days, in the leaves there is more production of growth active GA1 from its growth-inactive, precursor, GA20, due the increase in the enzyme that drives this step (GA 3-oxidase) (see pathway above and Figure 2). As a result there is less precursor GA20 left in the shoot.
There is a crucial difference between GA1 and GA20 in addition to their difference in activity in growth promotion: GA20 can move through the plant in the vascular system, whereas GA1 cannot do so.
In long days GA20 is not converted to GA1 in the leaves so there is more available to move to the stolon tips. In the stolon tips it is converted to GA1 and promotes stolon elongation. In short days the GA20 is converted to GA1 in the shoot, so there is little to be transported; the level of growth-active gibberellin in the stolon tips drops so enabling the tubers to form (see Figure 2).
To further demonstrate this mechanism Pratt made genetically modified potato plants over-expressing the genes for more total gibberellins or a higher amount of GA1 from GA20 in the shoot. In those with more GA1, tuber production occurred earlier than in those with more GA20.
We already indicated above that the genes encoding the production of the enzymes that regulate gibberellin biosynthesis are turned on and off. This itself is regulated by molecular switches, called transcription factors, that turn the genes on and off by binding to the DNA. From the work of David J. Hannapel and co-workers at Iowa State University, we know that there are several transcription factors actively involved in tuberization by regulating the production of hormones. These were discovered by analyzing which genes were being turned on in stolon tips. One of these encoded a transcription factor named POTH1. Plants genetically modified to over-express POTH1 are dwarf with enhanced tuber production. They have a reduced gibberellin biosynthetic pathway leading to a decrease in growth active gibberellin. Thus from numerous pieces of evidence we know that gibberellin is actively involved in promoting stolon elongation and inhibiting tuber formation, whereas tuberization is associated with a drop of gibberellins in the stolon tip.
Gibberellins are clearly not the entire answer. We know that the swelling of the tuber is associated with an increase in the content of a hormone class called cytokinins that are associated with cell division. These are also regulated by tuberization-associated transcription factors. Two years ago the flowering hormone was finally found after a 70 year search: it turned out to be a protein transcription factor whose synthesis is turned on in leaves and was then transported to the stem tip to induce flowering.
Potatoes contain an equivalent transcription factor that is synthesized in the leaves under short days and it, or its encoding signal (RNA), travels to the stolon where it mediates tuberization, possibly via cytokinins. At least part of the tuberization signal traveling from the leaves to the stolons is thus a transcription factor akin to the flowering hormone. Some day soon we will find out how all these signals interrelate.
Editor's Note: Dr. Peter J. Davies, a professor in the Departments of Plant Biology and Horticulture at Cornell University, has been chosen as the Potato Grower Researcher of the Year for 2008. This was a submission on Dr. Davies' research for potato tuberization and the signals that lead to it.